Abstract
Dynamic changes occurring in the lung microbiota can impact the initiation, progression, and prognosis of lung cancer (LC). Consequently, the development of suitable intratumoral microbiota analysis methods is crucial. Although matrix-assisted laser desorption/ionization mass spectrometry (MALDI MS) involves straightforward operations and provides precise results, the "direct smear method" limits the identification of bacterial subspecies. Furthermore, the issue of inadequate quantification with MALDI MS renders it unsuitable for direct analysis of intratumoral bacteria. To address these challenges, a novel ionic liquid in this study is employed, called norharmane conjugated to 2,5-dihydroxybenzoic acid (Nor@DHB) for the direct detection of intratumoral bacteria using MALDI MS. Because gram-negative bacteria are dominant within cancer cells, lipid A is selected as the chemical fingerprint for bacterial identification. The results demonstrated that using Nor@DHB can enhance the lipid A signal by an order of magnitude and achieved a good linear relationship within a concentration range of 0.01-80 ng mL(-1). Here, this method is successfully applied to the direct analysis of lipid A in actual clinical samples. Subsequent machine learning and nomogram models further confirmed the correlation between characteristic lipid A ions and LC patient clinicopathological features, which are further validated through both in vitro and in vivo experiments.